Photoionization Mass Spectrometric Study of the Prebiotic Species

Jan 19, 2010 - A photoion mass spectrometry study of the prebiotic species formamide was carried out using synchrotron radiation over the photon energ...
1 downloads 0 Views 254KB Size
J. Phys. Chem. A 2010, 114, 4847–4856

4847

Photoionization Mass Spectrometric Study of the Prebiotic Species Formamide in the 10-20 eV Photon Energy Range† Sydney Leach,*,‡ Hans-Werner Jochims,§ and Helmut Baumga¨rtel§ LERMA - UMR 8112, ObserVatoire de Paris-Meudon, 5, place Jules Janssen, 92195 Meudon, France and Institut fu¨r Physikalische und Theoretische Chemie der Freien UniVersita¨t Berlin, Takustr. 3, 14195 Berlin, Germany ReceiVed: October 22, 2009; ReVised Manuscript ReceiVed: December 16, 2009

A photoion mass spectrometry study of the prebiotic species formamide was carried out using synchrotron radiation over the photon energy range 10-20 eV. Photoion yield curves were measured for the parent ion and seven fragment ions. The ionization energy of formamide was determined as IE (12A′) ) 10.220 ( 0.005 eV, in agreement with a value obtained by high resolution photoelectron spectroscopy. The adiabatic energy of the first excited state of the ion, 12A′′, was revised to 10.55 eV. A comparison of the ionization energies of related formamides, amino acids, and polypeptides provides useful information on the varied effects of methylation and shows that polymerization does not substantially alter the ionization properties of the amino acid monomer units. Assignments of the fragment ions and the pathways of their formation by dissociative photoionization were made on the basis of ion appearance energies in conjunction with thermochemical data and the results of earlier electron impact mass spectral studies. Some of the dissociation pathways are considered to involve coupling between the 12A′ ground state and the low-lying 12A′′ excited state of the cation. Heats of formation are derived for all ions detected and are compared with literature values where they exist. Formation of the HNCO+ ion occurs by two separate paths, one involving H2 loss, the other H + H. In the conclusion a brief discussion is given of some astrophysical implications of these results. 1. Introduction Origin of life studies include the prebiotic formation of proteins from aminoacids and their precursors, as well as the formation of nucleobases that constitute an essential part of RNA and DNA structures. There have been numerous attempts to prepare these species from simpler compounds under conditions that resemble those of the primitive Earth, and there have been experimental and theoretical studies of analogous processes in the interstellar medium. UV and VUV radiation are among the important energy sources impinging on prebiotic and biotic species in astrophysical sites. Our previous VUV spectroscopy and photophysics studies on small prebiotic molecules,1 as well as on amino acids2 and purines and pyrimidines,3 including nucleobases,4 have been carried out in the context of their relevance to exobiological questions. The present study is on formamide, H2NCHO, a molecule that could be one of the precursors of these species.5 It is observed in the interstellar medium (ISM), in star-forming regions,6 and in the comet Hale-Bopp.7 Formamide, and its methyl derivatives could play a role in formation of adenine and related molecules in astrophysical milieux,5 and it has been suggested that formamide condensation occurring in the presence of inorganic oxides can be involved in a pathway of prebiotic synthesis of purine and pyrimidine derivatives. Indeed, it forms nucleobases in the presence of oxides when UV-irradiated8 or heated.9 †

Part of the special section “30th Free Radical Symposium”. * To whom correspondence should be addressed. Telephone: 33-1-45077561. Fax: 33-1-4507-7100. E-mail: [email protected]. ‡ Observatoire de Paris-Meudon. § Institut fu¨r Physikalische und Theoretische Chemie der Freien Universita¨t Berlin.

Formamide is the smallest and simplest model molecule of the peptide prototype NHsCdO linkage. This molecule, and its radical cation, can serve as models of the intact and oxidized peptide group. We remark that if indeed formamide is an acceptable model compound for the peptides, then the calculations of Robb and Csizamadia on its isomers10 support the suggestion11 that peptides could have been formed from nonamino acid sources in the primitive environment of the early Earth, in particular by direct synthesis from HCN polymer reactions with water without any intervening formation of R-aminoacids. It may be possible that such processes are occurring on Jupiter and, indeed, on other planets and on exoplanets whose atmospheres contain methane and ammonia, the precursors of hydrogen cyanide. Remarkably, peptides have been shown to replicate,12 which suggests that, preceding the RNA/DNA-based world, there could have existed a “protein world” in which peptides and proteins carry out all the essential biological functions (growth, selforganization, and evolution). The biochemical mechanisms involved would be expected to be relatively inefficient, compared with those active in RNA/DNA-based life. This relative inefficiency has been conjectured as a possible explanation of the long period (1.5 billion years) between the appearance of the first monocellular and the first multicellular organism.13 Our investigation in the 10-20 eV photon energy region is the first study of the dissociative photoionization of formamide. The results permit a comparison between ionization properties of some homopolypeptides and their related amino acids2 and formamides. We remark that the peptide bond is the sole covalent linkage between amino acids in the linear backbone structure of proteins. The amide functional group is also important as a repeat group in some industrial polymers, such as nylon.

10.1021/jp9098182  2010 American Chemical Society Published on Web 01/19/2010

4848

J. Phys. Chem. A, Vol. 114, No. 14, 2010

Leach et al.

TABLE 1: Adiabatic or Vertical (v) Energies (eV) of Formamide Ion States from High Resolution Photoelectron Spectroscopy, with Molecular Orbital Assignments19 and in Part from the Present Study cation ionization energy number

state energy and symmetry

molecular orbitals

IE 1 IE 2 IE 3 IE 4 IE 5 IE 6 IE 7 IE 8 IE 9

10.220a 12A′ 10.233b 10.55a 12A′′ 10.725b 13.686 22A′′ 14.607 22A′ 16.146 32A′ 18.8(v) 42A′ 20.5(v) 52A′ 29.2(v) 62A′ 33.1(v) 72A′

10σ a′ 2π a′′ 1π a′′ 9σ a′ 8σ a′ 7σ a′ 6σ a′ 5σ a′ 4σ a′

a Present study (see text). b Experimental values of ter Steege et al.18 Siegbahn et al.19 give 10.226 eV (12A′) and e10.699 eV (12A′′) for the adiabatic values.

2. Structural Aspects 2.1. Structure of Formamide (Neutral). Robb and Csizmadia10 discuss nine possible isomers of CH3NO stochiometry, divided into conformational, tautomeric, and structural isomers. Three tautomeric structures of formamide, H2NCHO, have been calculated by Bharatam et al.14 The lowest energy conformer is calculated to lie about 69-82 meV below the closest higher energy structure. Formamide can be considered to have two resonance structures that give the C-N bond partial double bond character with a planarized nitrogen atom.15 The planar molecule has Cs symmetry.16 The energy barrier to rotation about an amide bond is of fundamental importance in the modeling of protein conformation, but of course this can be affected by a solvent. We note that polypeptides can be considered as weakly coupled amide residues interacting through nonoverlapping charge distributions.17 2.2. Electronic Structure. The electron configuration of the planar Cs formamide molecule is

...(4a′)2(5a′)2(6a′)2(7a′)2(8a′)2(9a′)2(1a″)2(2a″)2(10a′)2 The a′ molecular orbitals (MOs) are σ orbitals, the a′′ are π orbitals. The two highest occupied orbitals are close together. Calculations of various degrees of sophistication, as well as diverging interpretations of the photoelectron spectra (PES) of formamide have led to order inversions, but the general consensus, well supported by the theoretical and experimental studies of ter Steege et al.,18 is that proposed by Siegbahn et al.19 in their high resolution PES study. This, and the corresponding ion state symmetries and energies, are given in Table 1. The experimental energy separation between the excited 12A′′ and the ground 12A′ state is 492 ( 28 meV according to ter Steege et al.19 (we find 330 meV, see later). Ruttink et al.20 calculated a value of 200 meV for the latter, Hop et al., 394 meV.21 Yu et al.22 calculated values that lie between -117 and 828 meV according to the level of calculation, and they conclude that in the 12A′ state the unpaired spin is primarily localized on the oxygen atom and the positive charge on the carbon atom. This is consistent with the C and N atoms being bonded with a double bond (calculated 1.297 Å) and that possibly a threeelectron bond exists between the C and O atoms (calculated 1.287 Å). In the 12A′′ state the charge and unpaired spin are delocalized over the C, N, and O atoms.

3. Experimental Section Synchrotron radiation from the Berlin electron storage ring BESSY II was monochromatized by a 3 m normal incidence monochromator and then focused into a differentially pumped gas cell. The general experimental setup is described in more detail elsewhere.23 Formamide vapor was introduced into the ionization chamber via a needle valve, with the sample at room temperature. Parent and fragment ions formed by photoionization were measured using a quadrupole mass spectrometer (Leybold Q200) (which tends to discriminate against higher-mass ions, especially at high mass resolution), and ion yield curves were obtained through photon energy scans with measuring intervals of 5 or 10 meV. The parent ions were all monomers and no dimers were detected. The yield curves of the principal ions observed are normalized to the incident photon flux measured by detecting the fluorescence of a sodium salicylate coated window. Wavelength-dependent photon flux changes are due to the grating transmission function and decreasing electron storage ring current. The yield curves are presented in appropriate figures. Spectral bandwidth of the incident monochromatic radiation was typically 0.2 nm. Ion appearance energies were determined mainly with the aid of semilog plots of the ion yield curves. Measurements were made up to 20 eV, but diminished grating reflectivity above 17 eV is responsible for enhanced noise in this spectral region, and we limit our discussed results to those below 18 eV. The formamide samples were commercial products (Sigma-Aldrich) of 98% stated purity. We remark that our ion appearance energies correspond to effective thermochemical energy values since they are a function of instrumental detection sensitivity and also reflect effects of intrinsic thermal energy as well as energy deposition in fragment products (eq 1). The kinetic energy shift has two main factors: limited detection sensitivity and the thermal energy stored in the parent neutral. Chupka has argued that compensatory effects lead to appearance energies that reflect reasonably well their 0 K values.24 Our AEs are determined from semilog ion yield plots in the threshold region by fitting straight lines to the noise and to the ion signal rise in this region. The photon energy at the intersection of these two lines is assigned to the measured AE value. Applying different fits, the precision is estimated by visual inspection of the variation of the intersection and is thus a function of the sharpness of the ion signal rise in threshold region. The measured AE values are used to calculate enthalpies of formation of fragment ions m1+ for different possible fragmentation pathways, using eq 2:

M + hν f M+ + e- f ml+ + mi AE + ∆fHgas(M) -

∑ [∆fHgas(mi)] ) ∆fHgas(ml+)

(1)

(2)

The ∆fH(m1+) values determined from eq 2 are then compared to tabulated standard thermochemical enthalpies of formation ∆fH(m1+), thus permitting assignment of particular fragmentation channels. If literature ∆fH(m1+) values are not available, then our values represent new, upper limit values of these entities. 4. Results and Discussion 4.1. Photoion Mass Spectrum. The mass spectrum, which was recorded at a photon excitation energy of 20 eV, is given

Photoionization Mass Spectrometric Study of Formamide

J. Phys. Chem. A, Vol. 114, No. 14, 2010 4849

TABLE 2: Photoion Mass Spectrum of Formamide Excited at 20 eV, Compared with Electron Impact Mass Spectra, As Well As Ion and Neutral Product Assignments and Photoion Appearance Energies [AE (eV)] m/z

photon impact

46 45 44 43

3 59 31 15

42 41 32 31 30 29 28 27 19 18 17 16 15

1 1 3 1 100 11 1 1 20 87 48 2

electron impact NIST 1 100 26 10

25

(Gilpin26) (100) (27) (12) (2) (0.3)

35 6 8 7 33 12 1

(1) (0.9) (28) (10) (12)

(1)

ion NH213CHO+ NH2CHO+ CH2NO+ HCNO+

neutral product

CNO+

H H2 H+H H2 + H

O2+a H3CO+? H-C-OH+? HCO+ HCNH+b HCN+? HNC+? H3O+?? H 2 O+ c NH3+ NH2+ NH+?

N? NH NH2 OH H2O? CN? HCN? CO HCO H2CO?

ion yield curves AE 10.220 ( 0.005 11.29 ( 0.01 15.59 ( 0.01 17.13 ( 0.02

13.11 ( 0.05 13.76 ( 0.02 12.6 11.37 ( 0.02 15.57 ( 0.02

a From air impurity. b N2+ occurs as an air impurity at AE ) 15.58 eV, see text. c Water impurity at 12.6 eV. H2O+ + HCN channel onset could occur at or above 13.446 eV, see text.

in Table 2, where it is compared with electron impact mass spectra listed by the NIST compilation25 and also by Gilpin.26 There is good agreement between the two reported electron impact mass spectra, but there are notable differences with the relative intensities of the m/z peaks of the photon-induced mass spectrum. In particular we note that the parent ion (m/z ) 45) is the most intense ion in electron impact whereas we find the m/z ) 29 ion to be the major ion in our photon impact measurement. In this respect it is interesting to note that in massresolved two- and three-photon resonance-enhanced ionization excitation spectra, ter Steege et al.18 also observed that the m/z ) 29 peak was the major peak. This indicates that our photon impact result is not due to a possible high-energy component of the photon flux. However, besides mass discrimination effects, due to differences in mass analyzers, there may be some other differences, between the photon impact and electron impact studies. There is the possibility for electron impact to form autoionizing triplet states, inaccessible by photon impact, which lead to enhanced parent ion formation. In addition, and in contrast to the absorption of photon energy, in electron impact ionization the approach of the electron can distort the neutral molecule before electron loss, thus making electron impact not a strictly vertical process.27 4.2. Parent Ion m/z ) 45. The yield curve of the parent ion, m/z ) 45 (Figure 1), provides a measurement of the IE of formamide as IE ) 10.220 ( 0.005 eV. This value will be discussed below and is compared with literature values, but first we discuss the structure of the cation and its excited electronic states. 4.2.1. Cation Structure. There have been several theoretical studies of the structure of the formamide cation. Bews and Glidewell28 have conceived of seven plausible structures: I [H2N-C(H)O]+; II [H2N-COH]+; III [HN-C(H)OH]+; IV [H3N-CO]+; V [HN-C-OH2]+; VI [N-C(H)OH2]+; VII [N-C-OH3]+. Their MINDO/3 calculations indicated that structure I has no bound solution but that structures II, IV, V, VI, and VII all have minima, with the order of increasing heat of formation being II < IV∼V < VI < VII. They conclude29 that the formamide cation can exist in five isomeric forms corresponding to structures II, IV, V, VI, and VII, with structure

Figure 1. Photoion yield curves of formamide: (a) parent ion, m/z ) 45, 10-11 eV; (b) parent ion, m/z ) 45, 10-18 eV.

III having no minimum but converging to V. However, MNDO calculations give a bound structure for I.30 This has been confirmed as the lowest energy stable form of formamide by more sophisticated calculations.20–22 A very thorough study of the isomeric species of stoichiometry [CH3NO]+ was carried out by Hop et al.21 They examined the ions in the gas phase by

4850

J. Phys. Chem. A, Vol. 114, No. 14, 2010

metastable ion, collisional activation, and charge stripping mass spectrometric experiments, including neutralization-reionization, to establish the stability of neutral structural analogues of the ions. Each of the parent ionic species (m/z ) 45) exhibited, in different abundances, fragments with m/z ) 44, 29, 17, and 16. Their calculations on the isomeric [CH3NO]+ ions indicate that they lie in deep potential wells. All barriers to rearrangement are predicted to be high, with the lowest being 1.94 eV for the conversion of H2NCHO+ to H2NCOH+ on the 12A′′ potential energy surface. We remark that the corresponding neutral isomers are also separated by high barriers.31 That there is negligible isomerization between ionized formamide and its tautomers formimidic acid H-NdC(H)-OH+ (III) and aminohydroxycarbene H2N-C¨ -OH+ (II) below the lowest ion dissociation energy limit has been demonstrated by experimental (CID and neutral-reionization) and theoretical studies.31 4.2.2. Isolated Electronic States? The question arises as to whether ion fragmentation is from isolated electronic states. This has been suggested, for example, by Bews and Glidewell,29 who conclude in their MINDO/3 theoretical study of formamide ion fragmentation study that H loss arises from electronic excited states. The formamide ion can exist as a σ (2A′) or a π (2A′′) radical (c.f., Table 1) and the chemistry of such radicals is expected to be different.21 However, to what extent they can act as isolated states merits discussion. As mentioned above, Hop et al.21 have shown, in reasonable agreement with experimental findings (Table 1) that, using RHF calculations with the GAMESS program (CI calculations), there is a stable excited state 12A′′ of the formamide ion at 394 meV above the 12A′ ground state. These are single reference calculations. Yu et al.22 point out that multireference calculations with additional electron correlation are required for systems that may have a number of low-lying electronic states. They carried out calculations at a significantly higher level than achieved by Hop et al.21 Their study was at various ab initio levels with basis sets up to 6-311+G(3DF,2P). Let us consider the case of no σ-π interaction. A σ radical cation would then be generated at 10.22 eV and a π radical cation at 10.55 eV. In the σ system, the charge is localized on the carbon atoms, whereas in the π system the charge is distributed over the O, C, and N atoms. However, the close proximity of these two electronic states favors their interaction so that the fragmentation of the ion at higher energies would not be from isolated states but from the ground state, of mixed σ and π parentage, in obeyance to the quasi-equilibrium theory of mass spectra.32 Dissociation channels, which become active above 11.29 eV (Table 2), will explore configurational space that has mixed percentage parentage of the σ and π states. For such a molecule as small as formamide, the possibility can exist of well-separated higher electronic excited states acting as isolated states. In that case, dissociation processes localized to these states, and even fluorescence, if the competitive timescales are favorable, could be observed from such excited states.33 The ≈3 eV energy gap between the 22A′′ and the 12A′′ ion states (Table 1) could possibly satisfy isolated states criteria33 and this would merit further attention. 4.2.3. Ionization Energy. In Table 3 we compare our measured value of the ionization energy of formamide with various published experimental values. Our adiabatic IE ) 10.220 ( 0.005 eV is in excellent agreement with that of the high resolution He I PES study of Siegbahn et al.,19 confirmed by the photoion and photoelectron study of ter Steege et al.18 It also agrees with the photoion spectroscopy value of Watanabe

Leach et al. TABLE 3: Adiabatic and Vertical Ionization Energies of Formamide technique PIMS (present study) 2-photon PI + PES18 high resolution He I PES19 He I PES45 He I PES46 He I PES47 He I PES48 He I PES49 He I PES50 He I PES51 He II PES39 XPS40 absorption spectroscopy36,37 photoionization34 photoionization35 electron impact mass spectrom.41 evaluation published data25

adiabatic IE (eV)

vertical IE (eV)

10.220 ( 0.005 10.42 ( 0.01 10.233 ( 0.008 10.419 ( 0.015 10.226 ( 0.005 10.423 ( 0.005 10.13 10.13 10.32 10.15 10.33 10.38 9.95 10.3 10.4 10.42 10.236 10.32 10.25 ( 0.02 10.16 ( 0.03 10.50 ( 0.05 10.16 ( 0.06

et al.34 but is just outside the limits of 10.16 ( 0.03 eV, the photoion spectroscopy value of Vilesov,35 although within the limits of the currently approved value of formamide, IE(ad) ) 10.16 ( 0.06 eV, in the NIST tabulation.25 Our IE(ad) also agrees with the early value, obtained by Hunt and Simpson36 by Rydberg state analysis, although the interpretation of the Rydberg features requires some modification.37 We note that a more recent analysis of Rydberg absorption features converging to the ground state of the formamide cation was carried out on the basis of IE(ad) ) 10.13 eV38 and may therefore need some revision or refinement in line with the Rydberg analysis study of ter Steege et al.18 There are some quite wild variations in the experimentally derived vertical IE values (Table 3). Our interpretation of steps in the parent ion yield curve (see below) gives a vertical IE value, 10.42 ( 0.01 eV, in agreement with that of Siegbahn et al.19 and with the values of He II PES39 and XPS40 studies. The reported electron impact value, 10.50 ( 0.05 eV,41 is presumably an onset value and is thus not necessarily the true vertical value, as it is actually higher than the photon impact vertical ionization energy. Concerning calculations of the IE of formamide, we note that those of Yu et al.22 at various levels give values of IE(ad) that range from 8.39 to 10.49 eV but that the vertical IE is calculated to be within the range 207-310 meV greater than the adiabatic value. This is in very reasonable agreement with the earlier experimental value 197 meV19 or the present study value 200 meV, which we consider to correspond to excitation of unresolved Duchinsky vibronically coupled ν4+ and ν5+ vibrational modes (Table 3 of ref 18). The geometry changes in going from the neutral to the cation are thus reasonably dealt with in the theoretical calculations of Yu et al. Their calculated adiabatic ionization energy that most closely approaches the experimental value 10.220 ( 0.005 eV is a G2(MP2) calculation, which gives IE(ad) ) 10.284 eV. It also gives IE(vert) ) 10.490 eV, which is 206 meV above the adiabatic value, in good agreement with experiment. 4.2.4. Comparison with Amino Acid and Polypeptide Ionization Energies. It is instructive to compare the ionization energy of formamide with that of films of aliphatic homopolypeptides. The latter have been measured by their photoelectron energy distribution spectra, for poly(glycine), poly(L-alanine), and poly(L-valine).42 The ionization energy values obtained are 7.3 ( 0.2 eV for poly(glycine) and poly(Lalanine), 7.2 ( 0.2 eV for poly(L-valine) (Table 4, column c).

Photoionization Mass Spectrometric Study of Formamide

J. Phys. Chem. A, Vol. 114, No. 14, 2010 4851

TABLE 4: Comparison of Ionization Energies (eV) of Related Formamides, Aminoacids, and Polypeptides (a) formamides (gas)

(b) aminoacids (gas)

(c) polypeptide (films)

(a) - (b)

(b) - (c)

(a) - (c)

formamide, 10.220 ( 0.005 methylformamide, 9.55 ( 0.04 dimethylformamide, 9.10 ( 0.02

glycine, 9.02 ( 0.02 R-alanine, 8.75 ( 0.05 R-valine, 8.9 ( 0.2

polyglycine, 7.3 ( 0.2 polyalanine, 7.3 ( 0.2 polyvaline, 7.2 ( 0.2

1.20 ( 0.025 0.80 ( 0.09 0.20 ( 0.22

1.72 ( 0.22 1.45 ( 0.25 1.70 ( 0.4

2.92 ( 0.21 2.25 ( 0.24 1.90 ( 0.22

Poly(glycine) is a simple peptide having peptide linkages, whereas poly(L-alanine) and poly(L-valine) have both peptide linkages and aliphatic units. The fact that the ionization energies are closely similar in all three peptides indicates that the ionization is of electrons from the peptide bonds, and thus bears comparison with the ionization of formamides. Furthermore, the quasi-constant values of the ionization energies of the three polymer films indicates that their ionization characteristics depend little on polymer conformation, since in the film measurements the main conformation of poly(L-alanine) is the R-helical structure, that of poly(L-valine) the β-pleated sheet, whereas poly(glycine) was considered to have a random structure partly consisting of two regular conformations, respectively, a pleated sheet and a 3-fold helix.42b Table 4 also includes a comparison between the ionization energies of formamides and aminoacids2 related to these polypeptides. The poly(glycine) peptide film values are about 3 eV below the formamide gas phase ionization energy, whereas the poly(L-alanine) and poly(L-valine) ionization energies are about 2.3 and 1.9 eV below those of N-methyl formamide43 and N,N′-dimethyl formamide,43 respectively (Table 4, (a) (c)). The data in column (a) shows that methyl groups attached to the nitrogen atom affect ionization in formamides whereas columns (b) and (c) have quasi-constant values, respectively, indicating that there is relatively little effect of methyl groups, attached to a carbon atom, on ionization in aminoacids and polypeptide films. We note that in formamide the nitrogen-carbon bond has some double bond character15 (subject to extensive discussion15b), thus allowing interaction with the pseudo-π methyl groups. When a methyl group is attached to the nitrogen atom in glycine, to form methylglycine (sarcosine), the ionization energy is lowered by about 600 meV,25 which is of the same order of magnitude as the difference, 670 meV, between the ionization energies of formamide and methylformamide. The quasi-constancy of the (b) - (c) values most probably reflects effects on ionization energies on going from gas to condensed phase (c.f., isobutene, about 2 eV44). Finally we remark that the (b) - (c) difference is constant within error limits, showing that polymerization does not substantially alter the ionization properties of the monomeric units. 4.2.5. Digression on the Double Ionization Energy of Formamide. It is of interest to remark that the double ionization energy of gaseous formamide has been determined to be 30 eV, obtained by subtracting the core hole binding energies from the Auger energies.52 In this determination, the Auger study value used averages of the Auger energies from the C, N, and O atoms in formamide, placed on a unique scale. The reported double ionization energy I2+ is 2.935 times the value of the single ionization energy I+. Tsai and Eland,53 on the basis of electron impact studies, have shown that for small molecules (diatomics to tetra-atomics) the average I2+/I+ ratio is 2.89, and that for aromatics it is 2.69, which is in good agreement with the ratio 2.65 ( 0.06 for PAHs and fullerenes in photon impact studies.54Thus, for formamide, the I2+/I+ ratio is close to the average for small molecules. Tsai and Eland proposed 2.8 ( 0.1 as a general value of the factor for all closed shell atoms and molecules. The I2+/I+ ratios can be rationalized53–55 on the

basis of Smith’s electrostatic model.56 We note that a MINDO/3 calculation of the double ionization energy gives a value of 23.12 eV for a structure considered to be H2NCOH2+,28 which is much below the reported experimental value,52 or the value expected from the Tsai and Eland I2+/I+ ratio, thus illustrating the inadequacy of the MINDO/3 calculation procedure which does not take into account a sufficient number of excitations or determine the adequate structure of the dication. 4.2.6. Heat of Formation of the Formamide Cation. The heat of formation of formamide is reported as ∆fH (H2NCHO) ) -1.93 eV.25 Using our value IE(ad) ) 10.22 eV, the heat of formation of the formamide cation becomes 8.29 eV. The listed value in the compilation of Lias et al.57 is given as 8.23 eV, based on a value of IE(ad) ) 10.16 eV. MNDO calculations give 8.40 eV,30 and higher theoretical level calculations give 8.24 eV.22 4.2.7. Structure in m/z ) 45 Ion Yield CurWe. Figures 1a and 1b shows that there are a number of steps or changes of slope in the parent ion yield curve. We divide these features into three sectors and report their measured energies and our assignments as follows, where IE 1, 2, 3,... correspond to ionic states enumerated in Table 1, and PES ionization features discussed are from Siegbahn et al.19 and ter Steege et al.18 i. Figure 1a. There are features at a number of photon energies, measured in first derivative plots of two separate yield curves of the parent ion. Their energies are noted here with their assignments: 10.220 ( 0.005 eV, assigned as IE1ad; 10.42 ( 0.01 eV assigned as IE1vert, the difference with IE1ad being 200 ( 15 meV, which could correspond to one quantum of the of the ν5+ vibration in the 12A′ ion state (201.7 meV18). The features at 10.55 ( 0.01 eV and 10.71 ( 0.01 eV require a more detailed discussion. We assign the 10.55 eV feature to the origin (adiabatic ionization energy) of the first excited ion state 12A′′ (IE2). There is a PES vibronic feature at essentially the same energy, at 10.56 eV, that we have measured from the well resolved PE spectrum published as Figure 5 of Siegbahn et al.,19 and there is also a two-photon ionization PES feature at 10.56 eV that we have determined from Figure 2 of ter Steege et al.18 This assignment of the adiabatic energy of 12A′′ differs from that of ter Steege et al.,18 who propose for IE2ad a peak in their PES at 10.725 ( 0.020 eV, corresponding to our ion yield feature at 10.71 eV. We reassign the 10.725 eV peak of ter Steege et al.18 to excitation of 2ν9+ in the 12A′′ state. This is the second member of a progression in ν9+, of which the first, the single ν9+ excitation, is overlapped by a strong peak at 10.419 ( 0.015 eV in the PES of ter Steege et al.18 The latter corresponds to our feature at 10.42 ( 0.01 eV in the ion yield curve which, as mentioned above, we have assigned as principally being the vertical IE1 of the 12A′ ion ground state. In Figure 5 of Siegbahn et al.19 we can pick out five members of the ν9+ progression in the 12A′′ excited state, from which we determine ν9+ ) 72 meV in this state (ν9 ) 70 meV in the ground state of neutral formamide). ii. Figure 1b. There are features at 12.0 eV; 12.58 eV; 13.45 eV; 14.11 eV, assigned to IE3vert ) 14.03 eV; and 14.68 eV (IE4ad ) 14.61 eV, IE4vert ) 14.75 eV). Uncertainty estimates are 10-20 meV for the photon energies. The first three features

4852

J. Phys. Chem. A, Vol. 114, No. 14, 2010

Figure 2. Photoion yield curves of formamide: (a) fragment ion m/z ) 44, 10-17 eV. In the inset the ordinate scale has been expanded by a factor of 25 over the energy range 10-12.5 eV; (b) fragment ion m/z ) 43, 14-18 eV; (c) fragment ion m/z ) 29, 12.5-17 eV.

of this series do not have any corresponding features in the photoelectron spectra19,39,45–51 but two of them do correspond to minima (11.95 eV) or maxima (13.33 eV) in the VUV absorption spectrum of formamide.58 Furthermore, Gingell et al.38 report EELS broad features at ∼11.7, ∼12.4, and ∼13.4 eV which appear to be similar to broad absorption bands: 11.15-11.95 eV, peak at 11.53 eV; 11.95-12.70 eV; 12.70-14.51 eV; and peak at 13.33 eV.58 These could correspond to our 12.0, 12.58, and 13.45 eV ion yield features. Gingell et al. make no definitive assignments of the three EELS bands but suggest that valence and Rydberg states are involved. We remark that the falloff in the parent ion yield curve beginning at 13.45 eV can

Leach et al. be related to the competitive dissociation channel leading to the major fragment ion m/z ) 29 whose AE ) 13.26 eV (see below). iii. The features in this third series can also be associated with PES ion state assignments (Table 1) or absorption features:58 Figure 1b: 15.91 eV (IE5ad ?); 16.27 eV (IE5vert ) 16.3 eV); and 17.28 eV (peak, which corresponds to an absorption peak at 17.2 eV58). 4.3. Fragment Ions. In this section we assign the fragment ions whose appearance energies were measured and discuss the corresponding dissociative ionization processes. The other ions in Table 2 are too weak for their yield curves to be measured. They are not discussed but, for completeness, their possible assignments are proposed in the Table. 4.3.1. m/z ) 44. The m/z ) 44 ion yield curve, given in Figure 2a, corresponds to the H loss channel. The fragment cation has the elemental composition NH2CO+. Five possible structures have been considered,28 of which the forms (a) H2N-C-O+ and (b) HN-COH+ but not (c) N-C-OH2+ can arise from the parent ion 12A′′ ground state dissociation, whereas (c) requires formation from an excited state such as 12A′. The slow character of the H-loss channel induced Bews and Glidewell29 to conclude that the m/z ) 44 ion arises from an electronically excited state. Experimental studies30,31 have shown that the loss of H is via a metastable parent ion where the departure of the carbon-bonded H produces the fragment ion H2N-CdO+, that is, structure (a), which is the most stable isomer of CH2NO+. A theoretical study of the unimolecular dissociation of the formamide ion was carried out by Ruttink et al.20 Their ab initio MO calculations at the MR-SDCI/CASSCF/DZPP +f level confirm that the loss of H is via a metastable parent ion losing the carbon-bonded H by direct cleavage. This reaction has a small reverse barrier, calculated to be 191 meV, in agreement with an experimental value of 130 meV.20 They also found that in the mass spectrum of deuterium-substituted formamide the D loss peak is absent, so that in formamide it is probable that H-loss from the ion occurs by tunnelling through the barrier. The appearance energy of the H2NCO+ ion is AE ) 11.29 ( 0.01 eV. Thus, the C-H bond dissociation energy in the cation is only about 1.07 eV. The reported electron impact appearance energy for m/z ) 44 is 12.00 eV,59 which is much higher than our photon impact value. The calculations of Ruttink et al.20 predict the appearance energy of H2NCO+ to be 1.03 eV above the ion ground state, in good agreement with our experimental value of 1.07 ( 0.015 eV. From our AE we calculate ∆fH(H2NCO+) ) 7.10 eV. This is smaller than the 7.26 eV value given in the Lias et al. compilation.57 Hop et al.60 found experimentally 6.96 eV, from the m/z ) 44 AEs for four different molecules (which do not include formamide). The corrected proton affinity derived value is given by Hop et al. as 7.04 eV.60 Our determination of ∆fH(H2NCO+) is the only one based on a photon impact investigation, whereas that of Hop et al.60 was by electron impact. We note that in the m/z ) 44 ion yield curve there are changes of slope at about 12.1, 12.65, 13.45 (a maximum as in m/z ) 45), 14.1, and 14.65 eV. These are similar to those of m/z ) 45 in the same energy range and can have the same assignments. We speculate that at higher energies the isomeric fragment ion, HNCHO+, could be formed by rupture of a N-H bond, assuming that its bond energy is greater than C-H bond energy, as appears to be the case for the neutral molecule,61,62 and as is consistent with the greater calculated length of the C-H bond

Photoionization Mass Spectrometric Study of Formamide as compared with N-H bondlengths, in the 12A′ and 12A′′ states of the formamide cation.20–22 4.3.2. m/z ) 43. This is the ion product of the H2 loss channel. The two possible fragment ions of m/z ) 43 are isocyanic acid HNCO+ and fulminic acid HCNO+, while little is known about a third possible isomer, cyanic acid HOCN+, which, in any case, we consider is highly unlikely to be formed. The ion yield curve can be considered to have two appearance energies at AE ) 15.59 ( 0.01 eV and at 17.13 ( 0.02 eV (Figure 2b). The lower energy AE corresponds to loss of molecular H2. From this we determine the heat of formation of the ion ∆fH(CHNO+) ) 13.66 eV. The values for the two possible CHNO+ ions in the Lias et al. compilation57 are ∆fH(HNCO+) ) 10.52 eV and ∆fH(HCNO+) ) 13.09 eV. The latter is closer to our experimental value, so we first consider that the m/z ) 43 ion is HCNO+. However, the Lias et al. value is an estimated value. If it is correct there is a barrier of 0.57 eV in the dissociation process. If the m/z ) 43 ion is indeed HCNO+, ours can be considered as an upper limit value of the heat of formation of HCNO+. Studies with a view of determining the extent of the barrier in such a reaction would then be welcome. We now examine whether the assignment to HCNO+ can be justified. The H2 loss process could be an elementary reaction if it occurs from the ground state of neutral formamide (involving rupture of a C-H and an N-H bond),61 or a twostep process, on the lines of that proposed by Liu et al.61 for elimination of H2 from the triplet state of neutral formamide: (i) N-H bond cleavage, forming the (triplet state in the neutral) complex H · · · NHCHO, and (ii) the departing H moves back toward the hydrogen atom of the COH unit. However, these dissociation processes would leave HNCO, not HCNO. Thus, rearrangement is required. It is thus instructive to consider whether isomerization of formamide could play a role, but inspection of formamide tautomers shows that there is no easy way of forming HCNO, and it is known that there is negligible isomerization between ionized formamide and its tautomers,31 at least at low excitation energies. Thus, it is possible that the dissociation process in the ion leads to HNCO+ rather than HCNO+, via a high energy barrier of about 3.14 eV. The corresponding barrier in neutral formamide is calculated to be about 3.69 eV.61,63 The second appearance energy, at 17.13 ( 0.02 eV, we consider as due to the loss of two hydrogen atoms: NH2CHO+ f HNCO+ + H + H. With this assignment we obtain a value of ∆fH(HNCO+) ) 10.68 eV, which is close to the Lias et al. compilation value ∆fH(HNCO+) ) 10.52 eV, thus confirming that the HNCO+ ion is formed in the dissociative ionization process. 4.3.3. m/z ) 29. The ion of m/z ) 29, whose yield curve is given in Figure 2c can have three possible assignments: (a) HCO+, (b) COH+, and (c) CH3N+. The first two would result from an NH2 loss channel, involving rupture of the C-N bond (although rearrangements could also be considered, see below), whereas (c) corresponds to loss of an oxygen atom from the parent ion. The appearance energy of the m/z ) 29 ion is AE ) 13.11 ( 0.05 eV, which is considerably smaller than the reported electron impact value, 13.70 eV.59 Our photon impact appearance energy value leads to determine that the heat of formation of HCO+ or of COH+ is ∆fH ) 10.71 eV. This is higher than the listed values ∆fH(HCO+) ) 8.56 eV, ∆fH(COH+) ) 9.98 eV.57 For the case of (c) we obtain ∆fH(c) )

J. Phys. Chem. A, Vol. 114, No. 14, 2010 4853 8.75 eV, which is much less than ∆fH(CH2dNH+) ) 11.3 eV, ∆fH(HCNH2+) ) 11.18 eV57 and so excludes both of the CH3N+ isomer ions. The most probable assignments for the m/z ) 29 ion are HCO+, with a barrier of 2.15 eV or COH+ with a barrier of 0.73 eV. We consider it more likely that the ion is HCO+, formed by direct cleavage to H-CdO+ + NH2 in agreement with assumptions in earlier studies.59 In this respect we note that metastable ion spectra have shown that there are virtually no rearrangement reactions leading to the most stable products of formamide ion dissociation. This is not the case in collisional activation studies where rearrangements out-compete the direct cleavage reaction in the formation of the m/z ) 29 ion,31 but these rearrangements are due to postcollisional isomerizations of energized ions. Two possibilities exist for the dissociation mechanism leading to HCO+ + NH2. It could be via the parent ion ground state since the charge is localized on the N atom in 12A′. The carbon-nitrogen bond rupture would then involve rotation about the CdN partial double bond, but this requires much energy. Another possibility arises from the ab initio calculations of Yu et al.,22 which suggest that NH2 (and NH2+) loss reactions can be explained by C-N dissociation from the 12A′′ state where the C-N bond is calculated to be essentially a single bond.20,22 The products would be in their ground electronic states in both cases. However, in another theoretical study, Ruttink et al.64 suggest that after crossing from the ion ground state into the excited 12A′′ π state, the stretching of the C-N bond does not lead immediately to dissociation to HCO+ + NH2; at 2.6 Å a fast and irreversible proton transfer occurs in the transient H2N · · · HCO+ complex, leading then to NH3+ + CO. This process will be discussed later. The appearance energy of HCO+ is about 1.74 eV above that of NH3+ (table 2), so we can consider the processes leading to HCO+ and to NH3+ as being in competition above the appearance energy of HCO+. The relative importance of these two processes will depend on the exploration of nuclear configurational phase space after excitation and would thus be modified with increasing excitation energy. 4.3.4. m/z ) 28. Although the m/z ) 28 yield curve (Figure 3a) includes features of the impurity N2+, observed by its characteristic ion yield curve structures above the ionization energy of nitrogen at 15.58 eV,65 there is an onset in the m/z ) 28 ion yield curve at a much lower energy, 13.76 ( 0.02 eV. We consider possible assignments of this onset to the ions CO+, H2CN+, and HNCH+. From thermochemical data25,57 the calculated AE values for these species are 14.31 eV (CO+), 11.51 eV (H2CN+), and 12.14 eV (HCNH+). This shows that m/z ) 28 is not due to CO+ and, furthermore, the absence of any new feature or inflection in the region between 13.76 and 15.56 eV, in particular at and above 14.31 eV, indicates that the reaction leading to CO+ + NH3 (the charge switch reaction with respect to NH3+ formation, see below) does not occur or that it requires passage over a considerable energy barrier. In addition, the absence of its characteristic ion yield structures at and above 14.01 eV65 shows that there is no sign of CO+ as an impurity. Thus, we can assign m/z ) 28 to either H2CN+ or HCNH+, with OH as neutral fragment. From our AE we calculate ∆fH(m/z ) 28) ) 11.43 eV. Lias et al.57 give ∆fH(HCNH+) ) 9.81 eV and ∆fH(CNH2+) ) 11.49 eV. Our AE thus supports the two possible assignments: (i) CNH2+ or (ii) HCNH+ with a barrier of 1.62 eV. HCNH+ is more probable since isomerization to the formimidic form via a 1,3 hydrogen shift, or some other suitable reorganization is required in order

4854

J. Phys. Chem. A, Vol. 114, No. 14, 2010

Figure 3. Photoion yield curves of formamide: (a) Fragment ion m/z ) 28, 13-17 eV. In the inset the ordinate scale has been expanded by a factor of 15 over the energy range 13-15 eV; (b) ion m/z ) 18, 12-16 eV; (c) fragment ion m/z ) 17, 10-17 eV; (d) fragment ion m/z ) 16, 15-18 eV.

to form OH and thus could involve an important energy barrier. The relative importance of this channel should be reduced by deuteration, as yet unstudied. The equivalent dissociation channel in neutral formamide has not been reported or been subject to theoretical study. 4.3.5. m/z ) 18. The ion signal at m/z ) 18 is certainly that of the H2O+ ion (Figure 3b). It shows the characteristic features of the ion yield spectrum of water, with an onset at 12.62 eV,65 so that at least a part of the signal is due to a water impurity. It behoves us to examine whether the formamide dissociation channel giving rise to H2O+ + HCN (HNC) could also be detected. From thermochemical data25,57 we estimate the thermochemical threshold energies as 13.446 eV for H2O+ + HCN,

Leach et al. and 14.128 eV for H2O+ + HNC. These values are well within the water impurity ion yield spectrum, so we are unable to affirm that these formamide dissociation channels are active. We note also that the water impurity did not arise from thermal dissociation of formamide into water and hydrogen cyanide since, in our 20 eV photon impact study (Table 2), the HCN+ signal (m/z ) 27) is very much weaker than the (m/z ) 18) H2O+ peak, yet HCN has an ionization energy, 13.6 eV,25 only 1 eV above that of water. The well-known hygroscopic nature of formamide is the most probable reason for the presence of the water impurity. 4.3.6. m/z ) 17. In assigning the m/z ) 17 ion, whose appearance energy is 11.37 ( 0.02 eV (Figure 3c), we can eliminate OH+ formed from a water impurity (expected AE ) 18.11 eV65) and OH+ resulting from a dissociation of the formamide cation to OH+ + H2CN, whose expected AE ) 18.28 eV. We assign the ion to NH3+, formed with CO loss. This dissociation channel of the formamide parent ion is competitive with the channel producing the m/z ) 29 fragment ion above 13.11 eV, thus accounting for the falloff in the m/z ) 17 ion yield curve beginning at about 13.4 eV. We note also the existence of features similar to those observed at 15.91 and 16.27 eV for the parent ion and that we have assigned as being associated with IE5. We also remark that the AE of the m/z ) 17 ion is higher than the ionization energy, 10.16 eV of ammonia,25 so that we can exclude that this ion results from thermal dissociation of formamide. From our AE we determine ∆fH(NH3+) ) 10.59 eV. The NIST value is 9.59 eV,25 so that NH3+ appearance has an energy barrier of 1 eV. Ruttink et al.20 calculate a barrier of 1.14 eV for the reaction forming NH3+ + CO and, based on the thermochemical data in Lias et al.,57 they predict 11.28 eV (which would be 11.39 eV using our formamide IE) as the appearance energy of NH3+, close to our experimental value of 11.37 eV. Our measurement thus validates the decarboxylation mechanism suggested by Ruttink et al.,20 who propose that CO loss is from the 12A′′ excited state of the cation. In the 12A′ ground state, in order for the σ structure [H2NdC(H)-O · ]+ to lose CO via a classical 1,2 hydrogen shift, the NH2 group would have to be rotated about the double bond, and this requires considerable energy. We recall that the 12A′ state has the leading electronic configuration (9a′)2 (2a′′)2 10a′. In this case the charge is localized on the N atom and the CdN bond favors the planar configuration, whereas the leading configurations of the 12A′′ state are (9a′)2 (2a”)2 3a′′,20 where the 3a′′ orbital is a π* MO, and (9a′)2 (2a′′) (10a′)218 with its singly occupied CN π M.O. These two leading configurations in the 12A′′ state result in the CN bond being essentially a single bond, so that the NH2 group may now more freely rotate, substantially lowering the effective barrier to dissociation. As mentioned above, after surface crossing 12A′ f 12A′′ the single C-N bond stretches, but at 2.6 Å a fast and irreversible proton transfer takes place in the transient H2N · · · · HCO+ complex (ion-dipole complex), leading to NH3+ + CO. We note that our experimental activation energy for CO loss is 1 eV, which is close to the calculated value of Ruttink et al.20 (“about 1.13 eV”) and is much lower than that calculated for neutral formamide, 3.25 eV.61 4.3.7. m/z ) 16. The appearance energy of the m/z ) 16 ion is 15.57 ( 0.02 eV (Figure 3d). This is much lower than the expected appearance energies of O+ from a water impurity (19.0 eV) or from O2 (18.73 or 17.28 eV, in addition there is negligible m/z ) 32 (O2+, Table 2)). Furthermore, the assignment to the O+ + CH3N dissociation channel is incompatible with thermochemical data on ∆fH(CH3N). We conclude that the m/z ) 16

Photoionization Mass Spectrometric Study of Formamide ion is NH2+ and that the dissociation has as products NH2+ + HCO. This corresponds to the charge switch reaction producing m/z ) 29 (HCO+). From our AE we obtain ∆fH(NH2+) ) 13.19 eV, in excellent agreement with the value listed in the NIST compilation,25 13.12 ( 0.06 eV. We can consider that dissociation is a simple CdN bond rupture process and that it occurs from the 12A′ state, where the charge resides on the N atom. In neutral formamide dissociation to NH2 + CHO occurs 3.92 eV above the neutral ground state.61 In the ion we find that the corresponding reaction NH2+ + CHO occurs at 5.37 eV above the ion ground state. But, as mentioned previously, Yu et al.,22 based on their ab initio calculations, suggest that NH2 loss (and NH2+ loss) reactions can be explained by C-N dissociation from the 12A′′ state where the C-N bond is essentially a single bond. The products would be in their ground electronic states in both cases. However, a dissociation energy of 5.37 eV is closer to that of a CN double bond than a single bond,66 which suggests that dissociation is indeed from the ground state of the formamide cation. Alternatively, the reaction could proceed from the quasisingle C-N bonded 12A′′ state but the HCO product being formed in an excited state, e.g. the HCO 2A′′ state at 1.15 eV above its ground state. Conclusion Our photoion mass spectrometry study of formamide was carried out using monochromatized synchrotron radiation over the photon energy range 10-20 eV. Photoion yield curves were measured for the parent ion and seven fragment ions. The latter involve processes of neutral loss of atomic and molecular hydrogen, CO, and OH, NH2, and HCO radicals, respectively. A possible dissociation channel giving rise to HCN as neutral product is also discussed. The ionization energy of formamide was determined as IE (12A′) ) 10.220 ( 0.005 eV, thus confirming the high resolution photoelectron value of Siegbahn et al.19 This σ ground state of the ion lies 330 ( 15 meV below the first excited π state, 12A′′, whose adiabatic energy was revised to 10.55 eV from analysis of features in the parent ion yield curve and in relevant photoelectron spectra. Determination of these ionization energies gives the possibility of choosing the best theoretical methodology adapted to the case of formamide calculations, as exemplified in the work of Yu et al.22 It should also lead to future revision of formamide Rydberg series analyses which have previously been made, in part, on the basis of less accurate ionization energies. Coupling between the 12A′ and 12A′′ states is invoked in discussions of the possible existence of isolated electronic states in the ion, and of the mechanisms of dissociative ionization forming, in particular, the fragment ions HCO+ and NH3+, which have related pathways. A comparison of the ionization energies of related formamides, amino acids and polypeptides provides information on the varied effects of methylation, found to be much more important for methyl groups attached to a nitrogen atom, as in the methylformamides, than for methyl radicals attached to a carbon atom in the aminoacids and polypeptides. Furthermore, polymerization hardly affects the ionization properties of the amino acid monomer units. Some consideration is given to the relation between the single and double ionization energies of formamide, discussed in terms of an electrostatic model. This may also have some application to astrophysics since molecular dications can be formed in the ISM67 and they are also actively studied in planetary atmospheres.68 The fragment ions were identified and the pathways of their formation were proposed on the basis of their appearance

J. Phys. Chem. A, Vol. 114, No. 14, 2010 4855 energies, aided by thermochemical data and the published results of electron impact mass spectral studies. Heats of formation are derived for all ions detected and are compared with literature values where they exist; the values determined are often new or revised with respect to previous publications. One result of particular interest involves the formation of the m/z ) 43 fragment ions for which arguments are given in favor of HNCO+ as product ion rather than HCNO+. Furthermore, the formation of the m/z ) 43 ion is shown to involve two separate channels, H2 and H + H loss, respectively, having different appearance energies. Another result is that the activation energies in dissociative ionization are smaller in H, CO and NH2 loss, and greater in H2 and HCO loss, processes than in related neutral photodissociation reactions of formamide. Finally, we remark that H Ly-R emission, which is important in the VUV in both the solar system and in the interstellar medium (ISM), has an energy of 10.2 eV which is just below but close to the ionization energy 10.220 ( 0.005 eV of formamide. Thus the ionization yield of formamide would be negligible in this onset region. We can estimate that at the HI limit 13.6 eV in the ISM the total ionization yield will be of the order of 37%, using a rule of thumb valid for many molecules.1d,69 The formamide ion is stable up to 11.29 eV (Hloss channel), that is, 2.3 eV below the HI limit at 13.6 eV, where the total ionization yield is estimated to be of the order of 25%. These results, as well as information on the formamide dication, should be integrated into models concerning the presence and possible survival of this prebiotic molecule in space. They particularly concern the observation of formamide in the ISM and in comets, where this species would be subject to UV and VUV irradiation. Acknowledgment. This paper is dedicated to the memory of Herb Broida and of Arnold Bass, pioneers who energized and contributed so much to the world of free radicals and their symposia. Support from the CNRS Groupe de Recherche “GDR Exobiologie” and from the CNRS Programme Interdisciplinaire “Origine des Plane`tes et de la Vie” is gratefully acknowledged. We thank BESSY, and in particular Dr Gerd Reichardt, for making available facilities and excellent technical and administrative support for our experimental studies, under EC contract I 3 RII 3-CT-2004-506008. Norbert Champion is thanked for help with formatting problems. References and Notes (1) (a) Formic Acid: Leach, S.; Schwell, M.; Dulieu, F.; Chotin, J. L.; Jochims, H.-W.; Baumga¨rtel, H. Phys. Chem. Chem. Phys. 2002, 4, 5025. Schwell, M.; Dulieu, F.; Jochims, H.-W.; Fillion, J.-H.; Lemaire, J.-L.; Baumga¨rtel, H.; Leach, S. J.Phys.Chem.A 2002, 106, 10908. (b) Ammonia: Leach, S.; Jochims, H.-W.; Baumga¨rtel, H. Phys. Chem. Chem. Phys. 2005, 7, 900. (c) Acetic acid: Leach, S.; Schwell, M.; Un, S.; Jochims, H.-W.; Baumga¨rtel, H. Chem. Phys. 2006, 321, 159. Leach, S.; Schwell, M.; Jochims, H.-W.; Baumga¨rtel, H. Chem. Phys. 2006, 321, 171. (d) Acetonitrile: Leach, S.; Schwell, M.; Un, S.; Jochims, H.-W.; Baumga¨rtel, H. Chem. Phys. 2008, 344, 147. Schwell, M.; Jochims, H.-W.; Baumga¨rtel, H.; Leach, S. Chem. Phys. 2008, 344, 164. (2) Jochims, H.-W.; Schwell, M.; Chotin, J. L.; Clemino, M.; Dulieu, F.; Baumga¨rtel, H.; Leach, S. Chem. Phys. 2004, 298, 279. (3) Schwell, M.; Jochims, H.-W.; Baumga¨rtel, H.; Leach, S. Chem. Phys. 2008, 353, 145. (4) Jochims, H.-W.; Schwell, M.; Baumga¨rtel, H.; Leach, S. Chem. Phys. 2005, 314, 263. (5) Saladino, R.; Crestini, C.; Costanzo, G.; Negri, R.; Di Mauro, E. Bioorg. Med. Chem. 2001, 9, 1249. (6) Rubin, R. H.; Swenson, G. W., Jr.; Benson, R. C.; Tigelaar, H. L.; Flygare, W. H. Astrophys. J. 1971, 169, L39. Hollis, J. M.; Lovas, F. J.; Remijan, A. J.; Jewell, P. R.; Ilyushin, Y. Y.; Kleiner, I. Astrophys. J. 2006, 643, L25. Nummelin, A.; Bergmann, P.; Hjalmarson, A.; Friberg, P.; Irvine, W. M.; Millar, T. J.; Ohishi, M.; Saito, S. Astrophys. J. Suppl. 2000, 128,

4856

J. Phys. Chem. A, Vol. 114, No. 14, 2010

213. Schilke, P.; Groesbeck, T. D.; Blake, G. A.; Phillips, T. G. Astrophys. J. Suppl. 2000, 108, 301. (7) Bockele´e-Morvan, D.; Lis, D. C.; Wink, J. E.; Despois, D.; Crovisier, J.; Bachiller, R.; Benford, D. J.; Biver, N.; Colom, P.; Davies, J. K.; Gérard, E.; Germain, B.; Houde, M.; Mehringer, D.; Moreno, R.; Paubert, G.; Phillips, T. G.; Rauer, H. Astron. Astrophys. 2000, 353, 1101. (8) Senanayake, S. D.; Idriss, H. PNAS. 2006, 103, 1194. (9) Costanzo, G.; Saladino, R.; Crestini, C.; Ciciriello, F.; Di Mauro, E. BMC EVol.Biol. 2007, 7, S1. (10) Robb, M. A.; Csizamadia, I. G. J. Chem. Phys. 1969, 50, 1818. (11) Csizamadia, I. G.; Csizamadia, V. M.; Koves, G. J. Physiol. Chem. Physics 1969, 1, 513. Matthews, C. N.; Moser, R. E. (a) Proc. Natl. Acad. Sci., U.S.A. 1967, 56, 1087. (b) Nature 1967, 215, 1230. (12) Lee, D. H.; Granja, J. R.; Martinez, J. A.; Severin, K.; Ghadiri, M. R. Nature 1996, 382, 525. (13) Rode, B. M.; Plankensteiner, K. Handbook of Biologically ActiVe Peptides; Kastin, A. J. Eds.; Academic Press: London, 2006; pp 14811486. (14) Bharatam, P. V.; Moudgil, R.; Kaur, D. J.Phys.Chem. A 2003, 107, 1627. (15) (a) Basch, H.; Hoz, S. Chem.Phys.Lett. 1998, 294, 117. (b) Kemnitz, C. R.; Loewen, M. J. J. Am. Chem. Soc. 2007, 129, 2521. (16) Hirota, E.; Sugisaki, R.; Nielsen, C. J.; Sørensen, O. J. Mol. Spectrosc. 1974, 49, 251. (17) Basch, H.; Robin, M. B.; Kuebler, N. A. J. Chem. Phys. 1968, 49, 5007. (18) ter Steege, D. H. A.; Lagrost, C.; Buma, W. J.; Leigh, D. A.; Zerbetto, F. J. Chem. Phys. 2002, 117, 8270. (19) Siegbahn, H.; Asplund, L.; Kelfve, P.; Hamrin, K.; Karlsson, L.; Siegbahn, K. J. Electron Spectrosc. Rel. Phen. 1974, 5, 1059. (20) Ruttink, P. J. A.; Burgers, P. C.; Terlouw, J. K. Int. J. Mass Spectrom. Ion, Proc. 1995, 145, 35. (21) Hop, C. E. C. A.; Chen, H.; Ruttink, P. J. A.; Holmes, J. L. Org. Mass Spectrom. 1991, 26, 679. (22) Yu, D.; Armstrong, D. A.; Rauk, A. Chem. Phys. 1996, 202, 243. (23) Schwell, M.; Dulieu, F.; Ge´e, C.; Jochims, H.-W.; Chotin, J.-L.; Baumga¨rtel, H.; Leach, S. Chem. Phys. 2000, 260, 261. (24) Chupka, W. A. J. Chem. Phys. 1959, 30, 191. (25) NIST Chemistry Webbook (June 2005), National Institute of Standards and Technology Reference Database. Available from http:// webbook.nist.gov (current 2009). (26) Gilpin, J. A. Anal. Chem. 1959, 31, 935. (27) Holmes, J. L.; Aubry, C.; Mayer, P. M. Assigning Structures to Ions in Mass Spectrometry; CRC Press: Boca Raton, 2007; p 19. (28) Bews, J. R.; Glidewell, C. J. Mol. Struct. 1980, 67, 141. (29) Bews, J. R.; Glidewell, C. J. Mol. Struct. 1980, 67, 151. (30) Kiplinger, J. P.; Maynard, A. T.; Bursey, M. M. Org. Mass Spectrom. 1987, 22, 534. (31) McGibbon, G. A.; Burgers, P. C.; Terlouw, J. K. Int. J. Mass Spectrom. Ion Proc. 1994, 136, 191. (32) Rosenstock, H. M. AdV. Mass Spectrom. 1968, 4, 523. (33) Leach, S.; Dujardin, G.; Taieb, G. J. Chim. Phys. 1980, 77, 705. (34) Watanabe, K.; Nakayama, T.; Mottl, J. J. Quant. Spectry. Rad. Transfer 1962, 2, 369. (35) Vilesov, F. I. SoV. Phys. Uspekhii. 1964, 6, 888. (36) Hunt, H. D.; Simpson, W. T. J. Am. Chem. Soc. 1953, 75, 4540. (37) Robin, M. B. Higher Excited States of Polyatomic Molecules; Academic Press: New York, 1975; Vol II, p 134.

Leach et al. (38) Gingell, J. M.; Mason, N. J.; Zhao, H.; Walker, I. C.; Siggel, M. R. F. Chem. Phys. 1997, 220, 191. (39) Åsbrink, L.; Svensson, A.; von Niessen, W.; Bieri, G. J. Electron Spectrosc. Rel. Phen. 1981, 24, 293. (40) Lisini, A.; Keane, M.; Lunell, S.; Correia, N.; Naves de Brito, A.; Svensson, S. Chem. Phys. 1993, 169, 379. (41) Baldwin, M. A.; Loudon, A. G.; Webb, K. S.; Cardnell, P. C. Org. Mass Spectrom. 1977, 12, 279. (42) (a) Awanami, S.; Oda, N. Radiat. Res. 1983, 95, 24. (b) Awanami, S.; Oda, N. Biopolymers 1980, 19, 1919. (43) Leach, S.; Jochims H.-W.; Baumga¨rtel, H. Chem.Phys., to be submitted. (44) Vermeil, C.; Matheson, M.; Leach, S.; Muller, F. J. Chim. Phys. 1964, 61, 596. (45) Turner, D. W.: Baker, C.; Baker, A. D.; Brundle, C. R. Molecular Photoelectron Spectroscopy; Wiley-Interscience: London, 1970. (46) Brundle, C. R.; Turner, D. W.; Robin, M. B.; Basch, H. Chem. Phys. Lett. 1969, 3, 292. (47) Kimura, K.; Katsumata, S.; Achiba, Y.; Yamazaki, T.; Iwata, S. Handbook of HeI Photoelectron Spectra of Fundamental Organic Compounds; Japan Sci.Soc.Press: Tokyo, 1981. (48) Kishimoto, N.; Osada, Y.; Ohno, K. J. Electron Spectrosc. Relat. Phenom. 2001, 114-116, 183. (49) Ballard, R. E.; Jones, J.; Sutherland, E. Chem. Phys. Lett. 1984, 112, 306. (50) Ballard, R. E.; Jones, J.; Read, D.; Inchley, A.; Cranmer, M. Chem. Phys. Lett. 1987, 137, 125. (51) Henriksen, L.; Isaksson, R.; Liljefors, T.; Sandstrom, J. Acta Chem. Scand., Ser. B 1981, 35, 489. (52) Correia, N.; Naves de Brito, A.; Keane, M. P.; Karlsson, L.; Svenson, S. J. Chem. Phys. 1991, 95, 5187. (53) Tsai, B. P.; Eland, J. H. D. Int.J. Mass Spectrom. Ion Phys. 1980, 36, 143. (54) Tobita, S.; Leach, S.; Jochims, H. W.; Ru¨hl, E.; Illenberger, E.; Baumga¨rtel, H. Can. J. Phys. 1994, 72, 1060. (55) Leach, S. Can. J. Phys. 2001, 79, 501. (56) Smith, F. T. J. Chem. Phys. 1961, 34, 793. (57) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, (Suppl. No. 1). (58) Leach, S.; Champion, N.; Jones, N.; Jochims H.-W.;Baumga¨rtel, H.; manuscript in preparation. (59) Loudon, A. G.; Webb, K. S. Org. Mass Spectrom. 1977, 12, 283. (60) Hop, C. E. C. A.; Holmes, J. L.; Ruttink, P. A.; Schaftenaar, G.; Terlouw, J. K. Chem. Phys. Lett. 1989, 156, 251. (61) Liu, D.; Fang, W.-H.; Fu, X.-Y. Chem. Phys. Lett. 2000, 318, 291. (62) Kang, T. Y.; Kim, H. L. Chem. Phys. Lett. 2006, 431, 24. (63) Martell, J. M.; Yu, H.; Goddard, J. D. Mol. Phys. 1997, 92, 497. (64) Ruttink, P. J. A.; Burgers, P. C. Org. Mass Spectrom. 1993, 28, 1087. (65) Berkowitz, J. Photoabsorption, Photoionization and Photoelectron Spectroscopy; Academic Press: New York, 1979. (66) (a) Cottrell, T. L. The Strengths of Chemical Bonds; Butterworths: London, 1954; (b) Stals, J. ReV.Pure Appl. Chem. 1970, 20, 1. (67) Leach, S. J. Electron Spectrosc. Relat. Phenom. 1986, 41, 327. (68) Gronoff, G.; Lilenstein, J.; Simon, C.; Witasse, O.; Thissen, R.; Dutuit, O.; Alcaraz, C. Astron. Astrophys. 2007, 465, 641. (69) Jochims, H.-W.; Baumga¨rtel, H.; Leach, S. Astron. Astrophys. 1996, 314, 1003.

JP9098182